| dc.description.abstract | The present study sought to evaluate trial designs and training designs that are commonly used in popular commercially available computer-assisted language-learning (CALL) programs. The first two experiments (Experiment 1a and 1b) compared the effects of passive viewing and active student response methods in vocabulary learning. Contingencies on active student responding were implemented in the form of compound matching to sample (MTS), in which each trial presented a compound sample (i.e., an auditory and a textual FL stimulus), to which the learner was required to respond by selecting a matching picture from an array of choices. This approach is characteristic of vocabulary-learning trials in Rosetta Stone®. Passive viewing was represented by stimulus pairing (SP) trials, in which the auditory FL stimulus, the textual FL stimulus, and the matching picture were presented simultaneously, and no response was required from the learner other than clicking to advance to the next trial. This approach is characteristic of vocabulary-learning trials in Memrise®. With respect to training design, the goal of Experiment 2 was to compare the efficiency with which participants acquired the same emergent stimulus relations as a result of an instruction that teaches a small number of relations (Compound MTS condition) versus an instruction that directly targets a greater number of relations separately (Mixed condition). The prediction for Experiments 1a and 1b was that the Mixed group would have an advantage over the two other conditions, as participants assigned to this condition received both response contingencies and reliable presentation of each picture along with its associated foreign auditory-visual compound. This prediction was not confirmed. By contrast, Experiment 2 confirmed the prediction that the compound MTS condition would require fewer trials to complete to mastery, while resulting in similar post- and follow-up test performance. | en_US |
